Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=ijmf20
The Journal of Maternal-Fetal & Neonatal Medicine
ISSN: 1476-7058 (Print) 1476-4954 (Online) Journal homepage: https://www.tandfonline.com/loi/ijmf20
One step versus two step approach for gestational
diabetes screening: systematic review and
meta-analysis of the randomized trials
Gabriele Saccone, Claudia Caissutti, Adeeb Khalifeh, Sara Meltzer, Christina
Scifres, Hyagriv N. Simhan, Sefa Kelekci, Osman Sevket & Vincenzo Berghella
To cite this article: Gabriele Saccone, Claudia Caissutti, Adeeb Khalifeh, Sara Meltzer, Christina Scifres, Hyagriv N. Simhan, Sefa Kelekci, Osman Sevket & Vincenzo Berghella (2019) One step versus two step approach for gestational diabetes screening: systematic review and meta-analysis of the randomized trials, The Journal of Maternal-Fetal & Neonatal Medicine, 32:9, 1547-1555, DOI: 10.1080/14767058.2017.1408068
To link to this article: https://doi.org/10.1080/14767058.2017.1408068
Accepted author version posted online: 20 Nov 2017.
Published online: 03 Dec 2017. Submit your article to this journal
Article views: 194
View Crossmark data
REVIEW
One step versus two step approach for gestational diabetes screening:
systematic review and meta-analysis of the randomized trials
Gabriele Sacconea , Claudia Caissuttib, Adeeb Khalifehc, Sara Meltzerd, Christina Scifrese, Hyagriv N. Simhanf, Sefa Kelekcig, Osman Sevkethand Vincenzo Berghellac
a
Department of Neuroscience, Reproductive Sciences and Dentistry, School of Medicine, University of Naples“Federico II”, Naples, Italy;bDepartment of Experimental Clinical and Medical Science, DISM, Clinic of Obstetrics and Gynecology, University of Udine,
Udine, Italy;cDepartment of Obstetrics and Gynecology, Division of Maternal Fetal Medicine, Sidney Kimmel Medical College of
Thomas Jefferson University, Philadelphia, PA, USA;dDivision of Endocrinology and Metabolism, Department of Medicine and
Obstetrics and Gynecology, Faculty of Medicine, McGill University, Montreal, Canada;eDepartment of Obstetrics and Gynecology,
Division of Maternal Fetal Medicine, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA;fDepartment of Obstetrics, Division of Maternal Fetal Medicine, Gynecology and Reproductive Sciences, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA;gDepartment of Obstetrics and Gynecology, School of Medicine, Izmir Katip Celebi University, Izmir, Turkey;
h
Department of Obstetrics and Gynecology, Bezmialem Vakif University, School of Medicine, Istanbul, Turkey
ABSTRACT
Introduction: To compare both the prevalence of gestational diabetes mellitus (GDM) as well as maternal and neonatal outcomes by either the one-step or the two-step approaches.
Material and methods: Electronic databases were searched from their inception until June 2017. We included all randomized controlled trials (RCTs) comparing the one-step with the two-step approaches for the screening and diagnosis of GDM. The primary outcome was the incidence of GDM.
Results: Three RCTs (n ¼ 2333 participants) were included in the meta-analysis. 910 were randomized to the one step approach (75 g, 2 hrs), and 1423 to the two step approach. No sig-nificant difference in the incidence of GDM was found comparing the one step versus the two step approaches (8.4 versus 4.3%; relative risk (RR) 1.64, 95%CI 0.77–3.48). Women screened with the one step approach had a significantly lower risk of preterm birth (PTB) (3.7 versus 7.6%; RR 0.49, 95%CI 0.27–0.88), cesarean delivery (16.3 versus 22.0%; RR 0.74, 95%CI 0.56–0.99), macroso-mia (2.9 versus 6.9%; RR 0.43, 95%CI 0.22–0.82), neonatal hypoglycemia (1.7 versus 4.5%; RR 0.38, 95%CI 0.16–0.90), and admission to neonatal intensive care unit (NICU) (4.4 versus 9.0%; RR 0.49, 95%CI 0.29–0.84), compared to those randomized to screening with the two step approach. Conclusions: The one and the two step approaches were not associated with a significant differ-ence in the inciddiffer-ence of GDM. However, the one step approach was associated with better maternal and perinatal outcomes.
ARTICLE HISTORY Received 13 October 2017 Revised 12 November 2017 Accepted 19 November 2017 KEYWORDS Diabetes; insulin; gestational diabetes mellitus; obesity
Introduction
Carbohydrates disorders, including gestational diabetes mellitus (GDM) and pregestational diabetes mellitus (DM), are common morbidities in pregnancy, with short-and long-term consequences to mothers, fetuses, short-and newborns. It has been estimated that about 6–18% of all pregnancies are complicated by DM in pregnancy [1–6]. The last report from the International Diabetes Federation (IDF) estimate that worldwide, approxi-mately 1 in 7 births in 2015 were complicated by some form of hyperglycemia during pregnancy [7].
Management for women with GDM includes diet, physical activity, oral hypoglycemic agents and/or insu-lin as needed [6,8]. The management of women with
GDM is aimed at achieving best possible glycemic con-trol, with normal or near normal glucose values, while avoiding hypoglycemia [4–6]. This management is effective in reducing maternal and neonatal morbidity and mortality [6,7]. Nevertheless, worldwide controversy exists regarding the best approach and criteria for GDM screening and diagnosis (Table 1) [6,7,15–25]. Several randomized controlled trials (RCTs), comparing the one step versus the two step approaches for GDM screening and diagnosis, have been published so far [22–24].
The aim of this systematic review and meta-analysis of RCTs was to compare the incidence of GDM, as well as the maternal and neonatal outcomes, by the one step versus the two step approaches.
CONTACTVincenzo Berghella vincenzo.berghella@jefferson.edu Department of Obstetrics and Gynecology, Division of Maternal Fetal Medicine, Thomas Jefferson University, 833 Chestnut Street, First Floor, Philadelphia, PA, USA
ß 2017 Informa UK Limited, trading as Taylor & Francis Group
2019, VOL. 32, NO. 9, 1547–1555
Materials and methods
This review was performed according to a protocol designed a priori and recommended for systematic review [26]. Electronic databases (i.e. Medline, Scopus, ClinicalTrials.gov, EMBASE, ScienceDirect, the Cochrane Library at the CENTRAL Register of Controlled Trials, SCIelo) were searched from their inception until January 2017. Search terms used were the following text words: “diabetes”, “trial”, “screening”, “diagnosis”, “one-step”, “two-step”, “guidelines”, “review”, “randomized”, and “clinical trial”. No restrictions for language or geographic location were applied. In add-ition, the reference lists of all identified articles were examined to identify studies not captured by elec-tronic searches. The elecelec-tronic search and the eligibil-ity of the studies were independently assessed by two authors (GS, CC). Differences were discussed with a third reviewer (VB).
We included all RCTs comparing the one step ver-sus the two step approaches for screening and diag-nosis of GDM. Quasi RCTs (i.e. trials in which allocation was done on the basis of a pseudorandom sequence, e.g. odd/even hospital number or date of birth, alter-nation) were excluded.
The risk of bias in each included study was assessed by using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions. Seven domains related to risk of bias were assessed in each included trial since there is evidence that these issues are associated with biased estimates of treat-ment effect: (1) random sequence generation; (2) allo-cation concealment; (3) blinding of participants and personnel; (4) blinding of outcome assessment; (5) incomplete outcome data; (6) selective reporting; and (7) other bias. Review authors’ judgements were cate-gorized as “low risk”, “high risk” or “unclear risk” of bias [26].
Two authors (GS, CC) independently assessed inclu-sion criteria, risk of bias and data extraction. Disagreements were resolved by discussion with a third reviewer (VB). All analyses were done using an intention-to-treat approach, evaluating women accord-ing to the screenaccord-ing group to which they were ran-domly allocated in the original trials. Primary and secondary outcomes were defined before data extrac-tion. All authors of the original trials were contacted for missing data.
The primary outcome was the incidence of GDM. Secondary outcomes were gestational weight gain (GWG) from randomization to delivery (in grams), ges-tational hypertension and preeclampsia (as defined by the original trial), preterm birth (PTB)< 37 weeks, induction of labor, shoulder dystocia (as defined by the original trial), cesarean delivery, and perinatal out-comes including birth weight, stillbirth (i.e. fetal death >23 weeks), macrosomia (i.e. birth weight 4000 grams), large for gestational age (LGA) (i.e. birthweight >90th percentile), small for gestational age (SGA) (i.e. birthweight <10th percentile), neonatal hypoglycemia (i.e. glucose <40 mg/dL), neonatal hyperbilirubinemia (i.e. total serum bilirubin>5 mg/dL), admission to neo-natal intensive care unit (NICU), and neoneo-natal death (i.e. death of a liveborn baby within the first 28 days of life). We also planned to assess cost-analysis com-paring the two screening methods.
The data analysis was completed independently by two authors (GS, CC) using Review Manager v. 5.3 (The Nordic Cochrane Centre, Cochrane Collaboration, 2014, Copenhagen, Denmark). The completed analyses were then compared, and any difference was resolved by discussion with a third reviewer (VB).
Data from each eligible study were extracted with-out modification of original data onto custom-made data collection forms. For continuous outcomes
Table 1. Criteria for gestational diabetes mellitus screening by selected societies. Test Number of abnormal values required for diagnosis Fasting glucose (mg/dL) 1 hour after loading (mg/dL) 2 hours after loading (mg/dL) 3 hours after loading (mg/dL)
ACOG, 2013 C&C [6] 2 step 3 h 100 g 2 95 180 155 140
ACOG, 2013 NDDG [6] 2 step 3 h 100 g 2 105 190 165 145
ADA, 2017 75g [9] 1 step 2 h 75 g 2 95 180 155 Not required
ADA, 2017 100g [9] 2 step 3 h 100 g 2 95 180 155 140
CDA, 2013 [10] 2 step 2 h 75 g 2 95 191 160 Not required
FIGO, 2013 [11] 1 step 2 h 75 g 1 92 180 153 Not required
IADPSG, 2015 [7] 1 step 2 h 75 g 1 92 180 153 Not required
NICE/RCOG, 2015 [12] 1 step 2 h 75 g 1 101 Not required 140 Not required
WHO, 1999 [13] 1 step 2 h 75 g 1 126 Not required 200 Not required
WHO, 2013 [14] 1 step 2 h 75 g 1 92 180 153 Not required
ACOG: American College of Obstetricians and Gynecologists; ADA: American Diabetes Association; CDA: Canadian Diabetes Association; C&C: Carpenter and Coustan; IADPSG: International Association of Diabetes Pregnancy Study Group; NICE: National Institute for Health and Care Excellence; RCOG: Royal College of Obstetricians and Gynaecologists; NDDG: National Diabetes Data Group; WHO: World Health Organization.
means ± standard deviation was extracted and imported into Review Manager v. 5.3.
Meta-analysis was performed using the random effects model of DerSimonian and Laird, to produce summary treatment effects in terms of mean difference (MD) or relative risk (RR) with 95% confidence interval (CI). Heterogeneity was measured using I-squared (Higgins I2). A subgroup analysis for the primary and the secondary outcomes was performed comparing the one step with 75 g 2 hours test using the International Association of Diabetes and Pregnancy Study Group (IADPSG) criteria, versus the two step with 50 g 1 hour followed for abnormals by the 3 h 100 g test using the Carpenter and Coustan (C&C) criteria.
Potential publication biases were assessed statistic-ally by using Begg’s and Egger’s tests. The meta-ana-lysis was reported following the Preferred Reporting Item for Systematic Reviews and Meta-analyses (PRISMA) statement [26]. Before data extraction was completed, the review was registered with the PROSPERO International Prospective Register of Systematic Reviews (CRD42017060752).
Results
Three RCTs [22–24] (n ¼ 2333 participants) were identi-fied as relevant and included in the meta-analysis (Figure 1, Table 2). Publication bias, assessed statistic-ally by using Begg’s and Egger’s tests, showed no sig-nificant bias (p ¼ .37 and p ¼ .31, respectively). All authors kindly provided additional unpublished data from their trial. The entire database from one trial was also obtained [23].
The overall risk of bias was low. All studies had low risk of bias in “random sequence generation”, and used opaque randomized envelopes. The randomiza-tion sequence was computer-generated by a statisti-cian. Adequate methods for allocation of women were used in all the trials. Given the intervention, no trial was double-blind (Figure 2). The statistical heterogen-eity within the study ranged from low to moderate with I2¼ 65% for the primary outcome, and I2¼ 0% for most of the secondary outcomes.
Out of the 2333 women included in the review, 910 were randomized to the one step approach (2 hrs, 75 g),
Records idenfied through database searching (n = 96) Sc re e ning Inc lude d Eligibility Iden fica o n
Addional records idenfied through other sources
(n = 0)
Records aer duplicates removed (n = 80)
Records screened (n = 80)
Records excluded on the basis of tle or abstract
(n = 74)
Full-text arcles assessed for eligibility
(n = 7)
Full-text arcles excluded, with reasons Ongoing trial (n=1) Prior meta-analysis (n=1) No RCT (n=2) Studies included in qualitave synthesis (n = 3) Studies included in quantave synthesis (meta-analysis) (n = 3)
Figure 1. Flow diagram of studies identified in the systematic review. (Prisma template [Preferred Reporting Item for Systematic Reviews and Meta-analyses]).
and 1423 to the two step approach (Table 1). Regarding the two step approach, one trial used 50 g 1 hr followed by 100 g 3 hrs [23]; in one trial all women had 50 g 1 hr test before randomization and were excluded if glucose 200 mg/dL and then women in the control group received 100 g 3 hrs [24]; finally, Meltzer et al. [22] was a three arms trial with two con-trol groups: two step 50 g 1 hour followed by 100 g 3 hrs, and two step 50 g 1 hour followed by 75 g 2 hours. For this review, both control groups of this trial [22] were considered as control group (Table 3).
Data about diabetes management were available for two RCTs. Both studies used four blood glucose measurements per day, fasting and 1 hour after main meals; glucose cutoff values were 95 mg/dL fasting and 140 mg/dL 1 hour after meals. Management
differed on cutoffs for change from diet to pharmaco-logic therapy, type of initial therapy, criteria for dose adjustment and criteria for modify therapy. Meltzer [22] gave precise indications on management, while Scifres [24] preferred changes based on clinician judgement (Table 4).
Tables 5and6show the primary and the secondary outcomes. No significant difference in the incidence of GDM was found comparing the one step versus the two step approach (8.4 versus 4.3%; RR 1.64, 95%CI 0.77–3.48; Figure 3). Women screened with the one step approach had a significantly lower risk of PTB, cesarean delivery, macrosomia, LGA, NICU admission, and neonatal hypoglycemia, compared to those randomized to the screening with the two step approach (Tables 5 and 6). Moreover, women in the
Table 2. Characteristics of the included trials. Location Population screened
Timing of screening
Risk factors for early screening
Fasting or not fasting state
at screening Sample sizea
Meltzer, 2010 [22]
Canada All pregnant women without pregestational DM
24–28 weeks Presence of multiple
risk factorsb Fasting 1500 (500 versus 1000)
Sevket, 2014 [23]
Turkey Singleton gestations without pregestational DM
24–28 weeks Not stated Fasting 786 (386 versus 400) Scifres,
2015 [24]
USA Spontaneous-conceived, single-ton gestations without pre-gestational DM
18–24 weeks Not stated Fasting 47 (24 versus 23)
aNumber in the one step versus number in the two step group. bAccording to the Canadian Diabetes Association 21.
DM: diabetes mellitus.
Figure 2. Assessment of risk of bias. (A) Summary of risk of bias for each trial; Plus sign: low risk of bias; minus sign: high risk of bias; question mark: unclear risk of bias. (B) Risk of bias graph about each risk of bias item presented as percentages across all included studies.
one step group had also a lower birth weight of about 135 grams (Table 6). Given that only one trial [22] compared the two screening approaches in terms of costs, pooled data for this outcome were not available. Subgroup analysis was performed for primary and secondary outcomes excluding Meltzer et al. [22], which was slightly different from the other two RCTs in terms of inclusion criteria (i.e. inclusion of also mul-tiple gestations) and in terms of GDM screening crite-ria (Table 2). The subgroup analysis for the primary outcome revealed that the one step approach with 75 g 2 hours test using the IADPSG criteria was associ-ated with a significant increase in the incidence of GDM compared to the two step approach with 50 g 1 hour followed for abnormals by the 3 h 100 g test using the C&C criteria (13.9 versus 5.7%; RR 2.43, 95%CI 1.54–3.82; Figure 4). This Meltzer trial [22] did not contribute data to quantitative meta-analysis for the secondary outcomes. Therefore, like in the main analysis, the one step approach with 75 g 2 hours test using the IADPSG criteria was associated with signifi-cantly lower risks of PTB, cesarean delivery, macroso-mia, LGA, NICU admission, and neonatal hypoglycemacroso-mia, as well as significantly lower birth weight compared to
the two step approach with 50 g 1 hour followed for abnormals by the 3 h 100 g test using the C&C criteria (Tables 5and6) [23,24].
Discussion
This meta-analysis from three RCTs showed that screening women with the one step approach was associated with a nonsignificant difference in the inci-dence of GDM, compared to the two step approach. However, our study provides evidence that the one step approach is associated with better maternal and perinatal outcomes, including significantly lower risks of PTB, cesarean delivery, macrosomia, LGA, neonatal hypoglycemia, admission to NICU, and lower mean birth weight. Moreover, even if not statistically signifi-cant, in many other secondary outcomes, which were probably underpowered as uncommon, we found a nonsignificant trend for benefit in the one step group (Tables 5 and 6). When comparing the one step with 75 g 2 hours test using the IADPSG criteria (as recom-mended currently by IADPSG [7], FIGO [11], and WHO [13]), versus the two step with 50 g 1 hour followed for abnormals by the 3 h 100 g test using C&C criteria
Table 3. Study design of the included trials.
Study group Study group cutoffs Control group (1) Control group cutoffs Control group (2) Meltzer, 2010 [22] One step
(2 hrs, 75 g) CDA: fasting 95 mg/dL; 1 h 190 mg/dL; 2 h 160 mg/dL Two step (50 g 1 hr; 100 g 3 hrs) NDDG: fasting 105 mg/dL; 1 h 190 mg/dL; 2 h 165 mg/dL; 3 h 145 mg/dL Two step (50 g 1 hr; 75 g 2 hrs) Sevket, 2014 [23] One step
(2 hrs, 75 g) IADPSG: fasting 92 mg/ dL; 1 h 180 mg/dL; 2 h 153 mg/dL Two step (50 g 1 hr; 100 g 3 hrs) C&C: fasting 95 mg/dL; 1 h 180 mg/dL; 2 h 155 mg/dL; 3 h 140 mg/dL – Scifres, 2015 [24]a One step
(2 hrs, 75 g) IADPSG: fasting 92 mg/ dL; 1 h 180 mg/dL; 2 h 153 mg/dL Two step (50 g 1 h; 100 g 3 h) C&C: fasting 95 mg/dL; 1 h 180 mg/dL; 2 h 155 mg/dL; 3 h 140 mg/dL –
aAll women in this study first had a 50 g 1 hr test before randomization and were excluded if glucose200 mg/dl.
CDA: Canadian Diabetes Association; NDDG: National Diabetes Data Group; IADPSG: International Association of Diabetes and Pregnancy Study Group; C&C: Carpenter and Coustan.
Table 4. Diabetes management and primary outcome.
Meltzer, 2010 [22] Sevket, 2014 [23] Scifres, 2015 [24] Management Frequency of glucose testing Prebreakfast and 1 h after meals
(QID)a Not stated 4x/day; fasting and 1 hr pp a
Glucose target values Fasting 75–95 1h PC meal 140a Not stated Fasting<95 mg/dL; 1 hr pp<140 mg/dLa
Cutoffs for change from diet to therapy
As above for 3 days in a row or 4/7 daysa
Not stated Per clinician judgementa
Type of initial therapy Lifestyle followed by insulin PRNa Not stated Glyburide or insulina
Dose and frequency of initial therapy
NPH HS 4–10 units or premeal 2–4
units to starta Not stated Per clinician judgement a
Criteria for pharmacologic therapy dose adjustment
Patients given adjustment algorithm
for q2d changes if not at targeta Not stated Per clinician judgement a
Criteria for adding or switch-ing pharmacologic therapy
Only switch was lifestyle to insulin if
not at targetsa Not stated Per clinician judgement a
Primary outcome Costs of screening and maternal and neonatal outcomes for overall study
Maternal and neonatal outcomes
Maternal and neonatal outcomes
a
Table 5. Maternal outcomes. GDM GWG (grams) Gestational hypertension PE PTB Shoulder dystocia Induction Cesarean delivery Meltzer, 2010 [ 22 ] 18/486 (3.6%) versus 36/982 (3.7%) Not stated Not stated Not stated Not stated Not stated Not stated Not stated Sevket, 2014 [ 23 ] 56/386 (14.5%) versus 24/400 (6.0%) Not stated 57/386 (14.8%) versus 60/400 (15.0%) a 5/386 (1.3%) versus 25/400 (6.3%) a 15/386 (3.9%) 32/400 (8.0%) a Not stated Not stated 65/386 (17.6%) 91/400 (22.8%) a Scifres, 2015 [ 24 ] 1/24 (4.3%) versus 0/23 13,244 ± 8391 versus 14,514 ± 8210 0/24 versus 0/23 a 1/24 (4.3%) versus 0/23 0/24 versus 0/23 a 1/24 (4.3%) versus 0/23 4/24 (18.2%) versus 6/23 (26.1%) 2/24 (8.7%) versus 2/23 (8.7%) Total 75/896 (8.4%) versus 60/1405 (4.3%) 13,244 versus 14,514 57/410 (13.9%) versus 60/423 (14.2%) 6/410 (1.5%) versus 25/423 (5.9%) 15/410 (3.7%) versus 32/423 (7.6%) 1/24 (4.3%) versus 0/23 4/24 (18.2%) versus 6/23 (26.1%) 67/410 (16.3%) versus 93/423 (22.0%) RR or MD (95% CI) 1.64 (0.77 –3.48) 1270 grams ( 6016 to 3476) 0.98 (0.70 –1.38) 0.49 (0.04 –5.60) 0.49 (0.27 –0.88) 2.88 (0.12 –67.29) 0.64 (0.21 –1.97) 0.74 (0.56 –0.99) I 2 65% Not applicable Not applicable 59% Not applicable Not applicable Not applicable 0% GMD: gestational diabetes mellitus; GWG: gestational weight gain; PE: preeclampsia; PTB: preterm birth; RR: relative risk; MD: mean difference; CI : confidence interval. Data are presented as number (percentage) or as mean difference ± standard deviation. Boldface data: statistically significant. aAdditional unpublished data kindly obtained by the original authors. Table 6. Perinatal outcomes. BW (grams) Stillbirth Macrosomia LGA SGA Neonatal hypoglycemia Neonatal hyperbilirubinemia NICU admission Neonatal death Meltzer, 2010 [ 22 ] Not stated Not stated Not stated Not stated Not stated Not stated Not stated Not stated Not stated Sevket, 2014 [ 23 ] 3,209 ± 613 versus 3,344 ± 522 a Not stated 11/386 (2.8%) versus 26/400 (6.5%) a 11/386 (2.8%) versus 26/400 (6.5%) a 11/386 (2.8%) versus 18/400 (4.5%) a 7/386 (1.8%) versus 19/400 (4.8%) a 24/386 (6.2%) versus 31/400 (7.8%) a 18/386 (4.7%) versus 38/400 (9.5) a 1/386 (0.3%) 4/400 (1.0%) a Scifres, 2015 [ 24 ] Not stated 0/24 versus 0/23 1/24 (4.3%) versus 3/23 (13.0%) 1/24 (4.2%) versus 3/23 (13.0) a 3/24 (12.5%) versus 3/23 (13.0%) a 0/24 versus 0/23 a Not stated 0/24 versus 0/23 a 0/24 versus 0/23 a Total Not stated 0/24 versus 0/23 12/410 (2.9%) versus 29/423 (6.9%) 12/410 (2.9%) versus 29/423 (6.9%) 14/410 (2.9%) versus 21/423 (5.0%) 7/410 (1.7%) versus 19/423 (4.5%) 24/386 (6.2%) versus 31/400 (7.8%) 18/410 (4.4%) versus 38/423 (9.0%) 1/410 (0.2%) versus 4/423 (0.9%) RR or MD (95% CI) 2 135 grams (2 214.73 to 2 55.27) Not applicable 0.43 (0.22 –0.82) 0.43 (0.22 –0.82) 0.69 (0.35 –1.33) 0.38 (0.16 –0.90) 0.80 (0.48 –1.34) 0.49 (0.29 –0.84) 0.26 (0.03 –2.31) I 2 Not applicable Not applicable 0% 0% 0% Not applicable Not applicable Not applicable Not applicable RR: relative risk; MD: mean difference; CI: confidence interval; BW: birth weight; LGA: large for gestational age; SGA: small for gestational age; NI CU: neonatal intensive care unit. Data are presented as number (percentage, number in the one step versus number in the two step groups). Boldface data: statistically significant. a Additional unpublished data kindly obtained by the original authors. 1552 G. SACCONE ET AL.
(as recommended currently by ACOG [6] and ADA [9]), the one step with IADPSG criteria was associated with significantly higher incidence of GDM (13.9 versus 5.7%), and significantly lower risk of PTB, cesarean delivery, macrosomia, LGA, neonatal hypoglycemia, admission to NICU, and lower mean birth weight (Tables 5and6).
This may be the first meta-analysis of RCTs compar-ing the one with the two step approach for screencompar-ing of GDM. A prior Cochrane review by Farrar et al. aimed to evaluate and compare different testing strategies for diagnosis of GDM to improve maternal and infant health while assessing their impact on healthcare ser-vice costs [27]. Six small RCTs, including 694 women, were analyzed. However, none of these included stud-ies compared one versus two step approach and therefore this analysis was not carried out [27].
Only three RCTs were included in our meta-analysis. The largest one [22], did not report data on maternal and perinatal outcomes and therefore the second larg-est trial [23], drives the statistics in terms of maternal and benefits for the one step approach. This is the major shortcoming of our meta-analysis. Meltzer et al. [22] was slightly different from the other two in terms of inclusion criteria (i.e. inclusion of also multiple ges-tations) and in terms of GDM screening criteria, and also had two different control groups (Table 3). Only one trial [22], compared the two screening methods in terms of costs. Meltzer et al. concluded that the two step approach was less expensive with equivalent diagnostic power compared to the one
step approach [22]. Another limitation of our study may be that lowering the diagnostic criteria in any way that increases the incidence of the disease will increase the false positive rate and include more patients who are on the normal side of the spectrum in the “diseased” category. And that will improve out-comes in the“diseased” group, because those patients aren’t really all “diseased”.
Different approaches and criteria for screening and diagnosis of GDM have been proposed in the literature (Table 1) [6,7,9–14]. The most common approaches are the one step and the two step ones. Recent contro-versy [6,7] has focused on the fact that IADPSG [7], FIGO [11], and WHO [13], as well as other societies, recommend the 75-gram 2 hours test using the IADPSG criteria, while ACOG [6] and ADA [9] recom-mend the two step with 50 g 1 hour followed for abnormals by the 3 h 100 g test using C&C criteria. The argument against the one step approach has been that it increases the incidence of GDM significantly, without proven improvement in maternal and/or peri-natal outcomes [6]. Our meta-analysis of RCTs revealed that indeed, compared to the two step with 50 g 1 hour followed for abnormals by the 3 h 100 g test using C&C criteria, the 75-gram 2 hours test using the IADPSG criteria is associated with significantly higher incidence of GDM (13.9 versus 5.7%), but also with sig-nificantly lower risk of PTB (3.7 versus 7.6%), cesarean delivery (16.3 versus 22.0%), macrosomia (2.9 versus 6.9%), neonatal hypoglycemia (1.7 versus 4.5%), admis-sion to NICU (4.4 versus 9.0%), and lower mean birth
Figure 3. Forest plot for the incidence of gestational diabetes mellitus. CI: confidence interval.
Figure 4. Forest plot for the incidence of gestational diabetes mellitus in subgroup analysis comparing the one step with 75 g 2 hours test using the International Association of Diabetes and Pregnancy Study Group (IADPSG) criteria, versus the two step with 50 g 1 hour followed for abnormals by the 3 h 100 g test using the Carpenter and Coustan (C&C) criteria. CI: confidence interval.
weight by 135 grams (Tables 5 and6). Clinically, these are important outcomes, which would advise towards recommending more globally the 75-gram 2 hours test using the IADPSG criteria, especially if confirmed by more level 1 data, and by cost-effectiveness analyses based on data from this meta-analysis. Data from our meta-analysis may also explain, perhaps in small part, the higher risks of some of these outcomes (e.g. PTB, cesarean delivery, macrosomia) in the USA, where the two step with 50 g 1 hour followed for abnormals by the 3 h 100 g test using C&C criteria is commonly used [6], compared to most European countries, where in general the 75-gram 2 hours test using the IADPSG cri-teria is more commonly used. The 75-gram 2 hours test using the IADPSG criteria has the added benefit of being only one step, and so giving potentially results for GDM diagnosis sooner, but the downside of requir-ing fastrequir-ing for all women berequir-ing screened for GDM.
In summary, the one and the two step approach screening were associated with nonsignificant differ-ence in the inciddiffer-ence of GDM. However, the one step approach was associated with better maternal and peri-natal outcomes, including lower risk of PTB, cesarean delivery, macrosomia, LGA, neonatal hypoglycemia, admission to NICU, and lower mean birth weight.
Disclosure statement
The authors report no conflict of interest.
Funding
No financial support was received for this study.
ORCID
Gabriele Saccone http://orcid.org/0000-0003-0078-2113
References
[1] Ashoush S, El-Said M, Fathi H, et al. Identification of metformin poor responders, requiring supplemental insulin, during randomization of metformin versus insulin for the control of gestational diabetes mellitus. J Obstet Gynaecol Res. 2016;42:640–647.
[2] Casey BM, Duryea EL, Abbassi-Ghanavati M, et al. Glyburide in women with mild gestational diabetes: a randomized controlled trial. Obstet Gynecol. 2015;126:303–309.
[3] Crowther CA, Hiller JE, Moss JR, et al. Effect of treat-ment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med. 2005;352:2477–2486. [4] Garner P, Okun N, Keely E, et al. A randomized
con-trolled trial of strict glycemic control and tertiary level
obstetric care versus routine obstetric care in the management of gestational diabetes: a pilot study. Am J Obstet Gynecol. 1997;177:190–195.
[5] George A, Mathews JE, Sam D, et al. Comparison of neonatal outcomes in women with gestational dia-betes with moderate hyperglycaemia on metformin or glibenclamide – a randomized controlled trial. Aust N Z J Obstet Gynaecol. 2015;55:47–52.
[6] Committee on Practice Bulletins—Obstetrics. Gestational diabetes mellitus. Practice bulletin No. 137. American College of Obstetricians and Gynecologists. Obstet Gynecol. 2013;122:406–416. [7] International Association of Diabetes and Pregnancy
Study Groups Consensus Panel, Metzger BE, Gabbe SG, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in preg-nancy. Diabetes Care. 2010;33:676–682.
[8] Caissutti C, Saccone G, Ciardulli A, et al. Very tight versus tight control: which should be the criteria for pharmacologic therapy dose adjustment in diabetes in pregnancy? Evidence from randomized controlled trials. Acta Obstet Gynecol Scand. 2017 [cited Nov 10]. DOI:10.1111/aogs.13257
[9] American Diabetes Association. 2. Classification and diagnosis of diabetes. Diabetes Care. 2017;40(Suppl 1): S11–S24.
[10] Canadian Diabetes Association Clinical Practice Guidelines Expert Committee. Diabetes and preg-nancy. Can J Diabetes. 2013;37:S168–SS183.
[11] Hod M, Kapur A, Sacks DA, et al. The International Federation of Gynecology and Obstetrics (FIGO) Initiative on gestational diabetes mellitus: a pragmatic guide for diagnosis, management, and care. Int J Gynecol Obstet. 2015;131(Suppl 3):S173–S211. [12] NICE guideline. Diabetes in pregnancy: management
from preconception to the postnatal period; 2015. [Internet]; Available from: https://www.nice.org.uk/ guidance/ng3/
[13] WHO. Definition, diagnosis and classification of dia-betes mellitus and its complications. part 1: Diagnosis and classification of diabetes mellitus; 1999. [Internet]; Available from: http://apps.who.int/iris/bitstream/ 10665/66040/1/WHO_NCD_NCS_99.2.pdf
[14] WHO. Diagnostic criteria and classification of hypergly-caemia first detected in pregnancy; 2013. [Internet]; Available from: http://apps.who.int/iris/bitstream/ 10665/85975/1/WHO_NMH_MND_13.2_eng.pdf
[15] Agarwal MM, Dhatt GS, Punnose J, et al. Gestational diabetes: dilemma caused by multiple international diagnostic criteria. Diabet Med. 2005;22:1731–1736. [16] Nicholson WK, Fleisher LA, Fox HE, et al. Screening for
gestational diabetes mellitus: a decision and cost-effectiveness analysis of four screening strategies. Diabetes Care. 2005;28:1482–1484.
[17] Meltzer S, Leiter L, Daneman D, et al. Clinical practice guidelines for the management of diabetes in Canada. CMAJ. 1998;159(Suppl8):S1–S29.
[18] National Institute for Health and Clinical Excellence (NICE). Diabetes in pregnancy NICE guidelines. London: NICE; 2008.
[19] Meltzer S, Leiter L, Daniman D, et al. Clinical practice guidelines for the management of diabetes in Canada. CMAJ. 1998;1998:S1–S29.
[20] Alberti KG, Zimmet PZ. Definition, diagnosis and clas-sification of diabetes mellitus and its complications. part 1: Diagnosis and classification of diabetes melli-tus provisional report of a WHO consultation. Diabet Med. 1998;15:539–553.
[21] The HAPO Study Cooperative Research Group, Metzger BE, Lowe LP, et al. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med. 2008; 358:1991–2002.
[22] Meltzer SJ, Snyder J, Penrod JR, et al. Gestational dia-betes mellitus screening and diagnosis: a prospective randomised controlled trial comparing costs of one-step and two-one-step methods. BJOG. 2010;117:407–415. [23] Sevket O, Ates S, Uysal O, et al. To evaluate the
preva-lence and clinical outcomes using a one-step method versus a two-step method to screen gestational
diabetes mellitus. J Matern Fetal Neonatal Med. 2014;27:36–41.
[24] Scifres CM, Abebe KZ, Jones KA, et al. Gestational dia-betes diagnostic methods (GD2M) pilot randomized trial. Matern Child Health J. 2015;19:1472–1480. [25] Caissutti C, Khalifeh A, Saccone G, et al. Are women
positive for the one step but negative for the two step screening tests for gestational diabetes at higher risk for adverse outcomes? Acta Obstet Gynecol Scand. 2017 [cited Nov 1]. DOI:10.1111/aogs.13254
[26] Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions, version 5.1.0 (update March 2011). The Cochrane Collaboration; 2011; [Internet]; [cited 2016 Jan 20]. Available from:
www.cochrane-handbook.org
[27] Farrar D, Duley L, Medley N, et al. Different strategies for diagnosing gestational diabetes to improve mater-nal and infant health. Cochrane Database Syst Rev. 2015;1:CD007122.
